1)The Role of State and Federal Government Despite expressions of support, state and federal governments cannot be relied upon as the primary financing source despite the precedent. As numerous agencies retain jurisdiction, in-kind requests for assistance with licensing and zoning may be more feasible. 2)The Role of Local and Regional Government Radar sites incur upfront and recurring costs to the landowner and it may be necessary for local governments to bear some of these. Since the benefits extend to surrounding communities, local and regional governments should be asked to facilitate cost-share negotiations amongst neighbors, and to serve as a portal to local industry. 3)Radar Siting Network geometry is a crucial consideration and should balance total coverage area with appropriate overlap. Pre-deployment negotiations with potential sites should be nearly simultaneous and one should have critical mass before generating any significant expenses. Proceeding asynchronously risks a significant loss of leverage to sites who will realize their importance to the overall network geometry. Siting tools to quickly generate alternate layouts are useful. 4)High Speed Internet Radar data is voluminous and timely delivery to customers is of the utmost importance. The presence of inexpensive high speed internet at or near a potential site is a big plus. The last hop may need to be wireless if trenching is overly costly or not an option. Piggybacking onto an existing network can result in significant cost savings, however bandwidth and security concerns will need to be allayed. Adaptive data compression and transmission techniques may be necessary if sufficient bandwidth cannot be obtained at a target price. 5)Computing Infrastructure Numerous algorithms are constantly at work generating enhanced products from raw radar data as well as from derivations thereof. These can require enormous CPU, memory, and storage resources, especially when the weather is most active. Dedicating sufficient machines to handle the worst case is expensive and most cycles will be spent idle. Hence the use of the compute cloud to acquire resources dependent on the state of the atmosphere results in substantial savings. 6)Products Tailored to Business A privately funded venture will have a shorter timescale over which to demonstrate value to stakeholders. Target users have widely varied levels of comprehension of radar data and work-flows in weather sensitive industries tend to be well established and tailored to current radar display and notification products. A baseline offering should be made available, but intense market research and two way interaction with customers is especially important in a field with high barriers to entry, and high ongoing operation costs. Agreements with value added resellers for distribution can be helpful, but it should not be assumed that data can be shoe-horned easily into existing products. If the third-party development effort is too great, in house development may be the only viable solution. 7)Maintenance Inevitably problems will occur that require on-site maintenance. Automated detection and classification of failure modes should be considered. The large majority of failures are simple and require little training to repair. For these, one should attempt to use local resources who can be briefed quickly and without formal training. More serious or unclassified problems are left to trained professionals. Siting designs that do not require ladder climbing greatly increases the pool of candidates for all types of repairs. 8)Automated Command and Control of Sensors The heterogeneous user needs of a necessarily diverse customer base require different sensor behavior at different times. Needs are bracketed with triggers, if X occurs, I need Y, or frequency, I need X every Y seconds. To satisfy the requests, CASA employed the concept of Distributed Collaborative Adaptive Sensing (DCAS), requiring an artificial intelligence backbone to diagnose the state of atmosphere and then to optimize radar behavior according to greatest user need. Non-linear optimization search space expansion with respect to complex weather scenarios requires on-the-fly repartitioning of radar node groupings under command and control protocols. Additionally, distributed chains of command, whereby local control is generated by each individual node, with the potential to be overridden by centralized control, is an effective strategy to promote graceful degradation in the event of networking problems. 9)Leveraging Existing Physical Infrastructure The line-of-sight requirement often implies construction of expensive towers and platforms, cranes for mounting, and potentially land acquisition. The expenses associated with physical deployment may be the singe most significant obstacle to a viable business model, even more so than the cost of the radar hardware. All efforts should be made to leverage existing infrastructure when possible. Building rooftops and cell towers will often already have existing physical security (fencing, surveillance, locks) , empty space to spare, power, and communications equipment already present. However, utilization of these resources may come with severe form factor restrictions on the radar for engineering safety. Flat panel, phased array radars may be the only feasible form factors to use with existing infrastructure. 10)Automated Operations A paradigm shift for a radar mesonet may be that individual radars may not need to be operational 24/7. Situational awareness algorithms taking cues from meteorological models and the Nexrad network can be useful to initiate or end operations. Automated calibration routines can be triggered at appropriate times. Additionally as severe weather often results in power grid disruption and backup power is a costly necessity, some radars may run partially or entirely on batteries or generators. Power management is a chief concern in such deployments. Removing the need for a manual operator reduces personnel costs, and timing operations appropriately conserves power, disk space, and hardware lifespan.
- Indicates paper has been withdrawn from meeting
- Indicates an Award Winner